https://downdetector.com/status/facebook/
Normally not worth mentioning random $service having an outage here, but this will undoubtedly generate a large volume of customer service calls.
Appears to be failure in DNS resolution.
yeah we are seeing same, appears that none of their NS are responding to dns queries… affecting all their properties, whatsapp, facebook, instagram etc
chris
https://downdetector.com/status/facebook/
Normally not worth mentioning random $service having an outage here, but this will undoubtedly generate a large volume of customer service calls.
Most probably, if not already.
Appears to be failure in DNS resolution.
Thanks for this Eric.
Also impacting Instagram and, apparently, WhatsApp.
Here’s a screenshot:
-mel beckman
same problem in Israel
And WhatsApp and Instagram. Twitter users nationwide agree anecdotally.
What I’m getting is DNS failure.
-George
Considering the massive impact of this it would be interesting to see some traffic graphs from ISPs that have PNIs with Facebook, or high volume peering sessions across an IX, showing traffic to FB falling off a cliff.
Looks like their auth DNS dropped out of the DFZ:
https://twitter.com/g_bonfiglio/status/1445056923309649926?s=20
https://twitter.com/g_bonfiglio/status/1445058771261313046?s=20
Not so much a DNS failure as an
“oops, where did the routes to our auth servers go?”
moment.
or, in this case, several moments… ^_^;;
Matt
Looks like they run there own nameservers and I see the soa records are even missing.
Curious if there is a malicious angle after the 60 Minutes story last night.
Looks bigger than DNS. Some people are saying they’ve disappeared off the DFZ.
Yeah it looks like there dns servers are just dead. I can’t get an response from them.
I see the same. Still see those prefixes via direct peering, but DFZ doesn’t have them. Their auths are not reachable even though I have routes for them.
Some other weirdness I’m poking a bit on that doesn’t seem like it could possibly be related to FB DNS though.
If you check your BGP routing tables, you’ll probably
find that it’s not so much the SOA records that are
missing, as it is the prefixes to reach the DNS servers
entirely.
I suspect the DNS entries on the servers themselves
may look fine from inside facebook, leading to a slower
diagnostic and repair, as it’s only from the outside world
the missing routing entries in the global table make the
problem so painfully visible.
Having the DNS team frantically checking their servers
may slow the resolution down, if it is indeed a BGP failure
rather than a DNS server failure situation, as it seems to
appear at the moment. ^_^;
Matt
In other news worker productivity is up 100% today.
-richey
Hanlon's razor.
"Never attribute to malice that which is adequately explained by stupidity"
Corollary - "But don't rule out malice."
They haven’t completely dropped off, but the big subnets certainly have. I’m only seeing 20-odd /24s from them via the DFZ, but everything larger still directly.